search

ohdsi-studies / PioneerWatchfulWaiting

This study is part of the joint PIONEER - EHDEN - OHDSI studyathon in March 2021, and aims to advance understanding of clinical management and outcomes of watchful waiting in prostate cancer.
Apache License 2.0
7 stars 18 forks source link

Generate cohort definitions from phenotype tracker #18

Closed MaximMoinat closed 3 years ago

MaximMoinat commented 3 years ago

Steps:

  1. Download phenotype tracker spreadsheet from MS Teams
  2. Run the python notebook parse_phenotype_tracker.ipynb to create cohort csv's
  3. Run insert_cohort_definitions.R to pull in json and sql cohort definitions from Atlas.

Note: this excludes all cohorts without an atlas link. At the time of writing this excludes cohorts O8, O9 and O13.

MaximMoinat commented 3 years ago

~Noticed that the new cohort numbering is not the same as the old. Working on a fix...~

MaximMoinat commented 3 years ago

@keesvanbochove ~The sql definitions pulled directly from Atlas look very different than the ones originally included. To be investigated.~ edit: after removing cohort stats, the sql's are quite similar. We can see that many concept_ids have been added in the mean time in the new Atlas cohorts though. ~The json definitions are formatted differently, but seem identical for most cohorts.~ edit: restored the original formatting

MaximMoinat commented 3 years ago

Fixes #17 , as the automatically generated cohort json definitions do not seem to contain the non-breaking spaces